[hpsdr] kordic, cordic

Georg Prinz getpri at t-online.de
Sat Apr 25 11:41:10 PDT 2009


Hello Michael,

as Kirk is on vacancies, I would like to ask you some questions on the
cordic/kordic code.
 
Going through the kordic-test-module:

- why the frequency-reg is not a signed variable?
- is it ok to interprete this frequency register as a phase-step, fo/fs?
- how can I be shure, that after every "@(posedge CLK_12MHZ)" a step of
30 degree will be done?

Going through the kordic-module:

- why the generate code for the stages starts at i=0? Stage 0 is defined
already!
- is it true that at line "Z[i+1] <= ...." the atan_table[i] is accessed
from 0 up to stage=19, only? What is the rest of the atan_table for? 

Going through the cordic-module_

this version is very tricky and is obviously done by very experienced
person! I need some more time to catch it up for the whole brilliancy!

But why all the rounding is done? Do we really need this inbetween
values?

Going through the cordig-lecture page 34, the constant An = 1.1647. I
think it shall be 1.647! The reason is, that the overall gain is
0.60725935 and the value for the constant to unify the equation shall be
1/0.60725935!

Best Regards

Vy73, Georg,dl2kp





More information about the Hpsdr mailing list